prompt-library-audit-and-rebuild

Automated workflow for auditing and mass-refining large-scale AI prompt libraries (JSON + Markdown) to ensure semantic character consistency and technical compliance.

Trigger Conditions

• After a structural rebuild of a prompt library where content integrity is suspect.
• When "generic" terms (e.g., 'a woman', 'a dog') are detected in production prompts.
• Before initiating high-volume ComfyUI batch runs to ensure character/gear anchors are present.

Workflow Steps

1. Audit Phase (Detection)

Use a recursive Python script to scan the library for:

• Character Anchor Compliance: Verify presence of specific names (e.g., 'Sienna', 'Aura') and unique physical descriptors (e.g., 'Golden-Rust hair', 'hazel eyes'). Flag any use of generic terms ('a person', 'the girl', etc.).
• Technical Fidelity: Confirm mandatory gear strings are present (e.g., 'Arri Alexa 65', 'Zeiss Master Prime').
• File Integrity: Ensure every .json payload has a corresponding, synchronized .md file with matching filenames.